Benchmarking the HP Z840 Workstation for Video, Part 2: Encoding

How Much of the Difference Related To SSD vs. HDD?

As discussed in Part 1 of this series, the Z840 used an SSD boot/system drive, so here I’m measuring the performance advantage gained from storing the content to be encoded on an SSD data drive over an HDD drive. As Table 3 (below) shows, for most encoding tools the difference in minor, though with Squeeze it’s significant.

Table 3. Rendering time differences between SSD and HDD drives in trials on the Z840. Click the image to see it at full size.

To explain, the two major groups show tests results performed with HTT enabled (on the left) and no HTT on the right. The green boxes in each section represent the fastest score within each section, the yellow the slowest. The Delta column shows the difference between the HDD drive and the fastest SSD drive. As you can see, the difference was negligible for all tests except those involving Sorenson Squeeze, which received a significant performance boost from the faster disk.

Intuitively, Squeeze’s enhanced performance on the faster drives relates to the increased file throughput; where other encoders produced six files at most (Vantage), Squeeze was encoding and writing 20 files to disk. Figure 2 (below) shows a CPU utilization graph on the Z840 when encoding with Squeeze from the three hard disks, first the Turbo SSD G2, then the hard disk, then the SATA-based SSD. As you can see, when reading and writing to the G2 drive, the Z840 achieved higher CPU utilization, particularly as compared to the HDD, which reduced encoding time.

Figure 2. CPU utilization with Squeeze while encoding from (and writing to) the listed disks.

Vantage is certainly capable of producing multiple output packages from a single input; in retrospect, I should have used a more aggressive encoding workflow. In your shop, if you’re producing multiple packages from a single input file, you might find some benefit in using SSD drives as the source and target. For most garden variety one-to-one encodes, or even one-to-six encodes, an SSD drive doesn’t appear to deliver substantial performance boosts.

HTT or Not HTT

What about whether to encode with HTT enabled or disabled? That differed by project and computer, and Table 4 (below) tells the tale. Specifically, the table shows the times for all tests with HTT enabled (on the right) and disabled (on the left). The faster scores are in green, the slower scores in yellow. The predominance of green on the right tells you that HTT is beneficial in 17 out of 20 test cases.

Table 4. Performance differences with HTT disabled (on the left) and enabled (on the right). Click the image to see it at full size.

Were there any instances where you might consider running without HTT? Table 5 (below) compares the times, and posts negative numbers where HTT slowed encoding. None reached the 15% threshold that triggered a red background. In contrast, there are four instances where enabling HTT produced performance increases of 20% or more, which are those tests with a green background. At least for these encoding tools, running with HTT enabled is definitely advised.

Table 5. Comparisons of tests with and without HTT enabled. Negative numbers indicate that performance was faster with HTT disabled.

So, here are the key takeaways:

• A newer system like the Z840 should significantly accelerate encoding performance across a range of project types and formats.
• Using an SSD drive to store content to be encoded may deliver substantial reductions in rendering times, but only with encoders with very high file output, meaning that a single file is rendered into 20 or more iterations.
• With encoding tools, enabling HTT will almost always produce a faster result, though your mileage may vary.

This ends the encoding section; in the next segment we’ll look at analysis.